|
(11) | EP 0 135 243 B1 |
| (12) | EUROPEAN PATENT SPECIFICATION |
|
|
| (54) |
A method of producing a semiconductor structure on a substrate and a semiconductor device manufactured thereby Verfahren zum Herstellen einer Halbleiterstruktur auf einem Substrat und danach hergestellte Halbleiteranordnung Un procédé de fabrication d'une structure à semi-conducteur sur un substrat et dispositif à semi-conducteur fabriqué selon ce procédé |
|
|
|||||||||||||||||||||||||||||||
| Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention). |
[0001] This invention is concerned with a method of producing a semiconductor structure on a substrate having doping of a first type.
[0002] Fabricating integrated circuits with less coupling capacitance and higher density is a major goal in the manufacturing of very large scale inegrated (VLSI) circuits. As device geometries are reduced to the micrometer and submicrome- ter range, further developments in silicon gate metal-oxide semiconductor (MOS) processing methods are needed.
[0003] In the prior art, CMOS structures having a common gate for both n and p channel devices have been fabricated using bulk silicon for the p channel device (lower transistor) and a laser recystallized silicon film for the n channel device (upper transistor), (see e.g., J. F. Gibbons and K. F. Lee, "One-Gate Wide CMOS Inverter on Laser-Recrystallized Polysilicon", IEEE Electron Device Letters, Vol. EDL-1, No. 6, June, 1980). The Gibbons and Lee structure is vertically built and obtains high packing density by means of a self aligned common gate structure. However, the CMOS process used to produce this common gate device results in a complete source and drain overlap of the gate for the upper transistor. This overlap contributes to large coupling capacitance, which is undesirable for high performance MOS devices. Reduction of the coupling capacitance results in high circuit speed and low power drain in the device. Thus, it is important to obtain a MOS process which minimizes coupling capacitance and obtains high packing density. In addition, it is desirable that such a process be compatible with standard VLSI processing.
[0004] The present invention provides a method of producing a semiconductor structure on a substrate having doping of a first type and comprising the steps as specified in claim 1.
[0005] A method as set forth in claim 1 may further comprise the steps of depositing a first conductive material in selected active regions of said semiconductor substrate, covering selected regions of said first layer with a first mask, removing said first layer not covered by said first mask over at least one of said active regions to form an opening through said first layer, thereby exposing said semiconductor substrate surface, removing said first mask, and forming a second conductive material in said opening through said first intermediate insulating layer.
[0006] A method as set forth in the last preceding paragraph may further comprise the steps of farming a layer, comprising insulating material, - on the surface of said fifth layer, covering selectively a region of said layer material with a second mask, removing said layer not covered by said second mask to form an opening through said third intermediate insulating layer, removing said second mask, and forming a third conductive material in said second opening.
[0007] In carrying out a method as set forth in the last preceding paragraph, it is preferred that said third conductive material comprises a metal.
[0008] The present invention also provides a method of producing a semiconductor device on a substrate having doping of a first type, comprising the steps specified in claim 5.
[0009] By the method of the present invention a semiconductor device may be made on a substrate comprising first vertically stacked dual MOS transistor devices of a first type having a first common gate region, second vertically stacked dual MOS transistor devices of the first type having a second common gate region, and a plurality of source and drain regions coupled to the first and second common gate regions.
[0010] In a device as set forth in the last preceding paragraph it is preferred that said first vertically stacked dual MOS transistors are both enhancement mode, and said'second vertically stacked dual MOS transistors are both depletion mode.
[0011] The present invention solves the problems of the prior art by preventing source and drain overlap of the gate for the upper transistor. The invention includes new MOS processing sequences for producing a vertically built MOS structure in either CMOS or NMOS technologies which permits the out diffusion of dopant from a layer of chemically vapor deposited (CVD) doped oxide (for example phosphosilicate glass) into a layer of CVD laser recrystallized polysilicon. This out-diffusion is usually accomplished by subjecting an intermediate structure to a high temperature oxidation treatment. Subsequent source and drain mask alignment is chosen such that this out diffusion of dopant from the CVD glass at its boundary limit will meet with a diffusion of implanted source and drain ions.
[0012] A layer of thermally grown stress relief silicon dioxide (SRO) and a low pressure chemically vapor deposited (LPCVD) silicon nitrode layer are used above a selected area of laser recrystallized polysilicon as a mask against subsequent oxidation. During this oxidation which is done at high temperature, the dopant out diffuses from the doped oxide thereby doping the adjacent laser recrystallized polysilicon region. The source and drain zones are then defined by anisotropic etching of the SRO and silicon nitride. The mask alignment is selected to cover the boundary limits of dopant into the laser recrystallized polysilicon. In addition to these out diffused regions, additional diffusion of the source and drain implants occur during high temperature anneal and drive-in.
[0013] The mask alignment adjustment in this sequence makes possible minimal overlap of the drain and source zones with the gate, thus reducing coupling capacitance while providing increased packing density due to the vertically built structure. In addition, this process sequence requires only conventional VLSI fabrication techniques.
[0014] There now follows a detailed description, which is to be read with reference to the accompanying drawings, of several methods according to the invention and of semiconductor devices produced thereby; it is to be clearly understood that these methods and devices have been selected for description to illustrate the invention by way of example and not by way of limitation.
[0015] In the accompanying drawings:
Figures 1A to 1 N illustrate a first embodiment of a self-aligned vertically built common gate MOS structure and method according to the present invention;
Figure 2 is a schematic diagram of the embodiment illustrated in Figure 1;
Figure 3 illustrates an alternative structure to that shown in Figure 1;
Figure 4 illustrates a second embodiment of a selfaligned vertically built common gate MOS structure according to the present invention;
Figure 5 illustrates a third embodiment of a self- aligned common gate MOS structure according to the present invention;
Figure 6 illustrates a fourth embodiment of a self-aligned vertically built MOS structure according to the present invention; and
Figure 7 is a schematic diagram of the embodiment illustrated in Figure 6.
[0016] Figures 1Ato 1 N illustrate a first embodiment of the present invention. In Figure 1A on a silicon substrate 100 typically doped with n- dopant, a stress relief oxide (SRO) layer 130 is formed, typically consisting of thermally grown silicon dioxide (SRO). A first masking layer 120, e.g., a low pressure chemically vapor deposited (LPCVD) silicon nitride film is formed over the first SRO layer 130. A photoresist layer 110 on top of the first nitride layer 120 is utilized so that the first nitride layer 120 can be removed over a region 101 of the substrate 100 by etching for example by a fluorocarbon plasma. A trench 103 is formed by complete removal of nitride layer 120 in the region 101. The first nitride 120, first SRO 130, and photoresist 110 layers form a mesa 104 above the trench 103 with no overhang. A channel stop (e.g., n+ dopant) can then be implanted in the trench 103 if desired.
[0017] In Figure 1B the mesa 104 is shown overlain by the first nitride 120 and first SRO 130 layers after the photoresist mask 110 is removed in preparation for further masking. The first SRO layer 130 remains in the trench 103.
[0018] Figure 1 C shows growth of a first field oxidation (FOX) layer 140, over the first SRO layer 130, located in the trench 103. It should be noted that in practice the first SRO 130 and the first FOX 140 layers, although shown as separate entities for clarity, do not remain separate but merge together since they are chemically of the same material.
[0019] In Figure 1 D the first nitride layer 120 and the first SRO layer 130 over the mesa region 104 are removed to expose the substrate 100.
[0020] A first gate oxidation (GOX) layer 150, as shown in Figure 1 E, is thermally grown over the exposed surface of the substrate 100. If desired, a gate threshold control implant can then be applied in the first GOX layer 150. Then a first polysilicon (POLY) layer 160 is deposited over the entire surface of the substrate 100 covering the first SRO layer 130, the first FOX layer 140, and the first GOX layer 150. The first POLY layer 160 is formed typically by low pressure chemically vapor deposited (LPCVD) techniques. The entire first POLY layer 160 is subsequently doped for example p+ with boron.
[0021] A positive photoresist layer 162 is then used to define regions 164 as shown in Figure 1 and the first POLY layer 160 is removed over the rest of the surface 165 by etching. Thus trenches 167 and 168 are defined. Source 170 and drain 171 ion implants (e.g., with p+ material) followed by a high temperature drive-in are then made in the trenches 167 and 168 respectively as shown in Figure 1G. This implanting sequence also dopes the POLY layer 160.
[0022] After the photomask 162 is removed a first doped glass layer 172 (e.g., silicon dioxide doped with n+ phosphorus) is deposited as shown in Figure 1 H over the entire surface 165 and over the first POLY layer 160.
[0023] The doped glass layer 172 is subsequently anisotropically etched using a directional plasma as shown in Figure 11. After the anisotropic etching, two small regions 173 and 174 of doped glass will remain adjacent to the first POLY region 160. A second gate oxidation (GOX) layer 180 is then thermally grown over the entire surface 165 and the POLY region 160 and etched leaving the second GOX layer 180 above the first POLY region 160 only.
[0024] A second polysilicon (POLY) layer 190 as shown in Figure 1J is deposited over the entire surface 165, the small regions 173 and 174, and the second GOX 180. The second POLY layer 190 is formed typically by Ipcvd techniques and is subsequently doped for example with p- material. The second POLY layer 190 is then recrystallized, for example by use of a laser.
[0025] As illustrated in Figure 1K a second SRO layer 200 is grown over the laser recrystallized POLY layer 190. Then a second masking layer 210, e.g., LPCVD silicon nitride, is deposited over the second SRO layer 200 and a positive photoresist layer 220 is used over the layer 210 to define an active region 230. The second nitride layer 210 and the second SRO layer 200 are then revoved over a region 240 to expose the laser recrystallized POLY layer 190. After the photoresist layer 220 is removed over region 230 a second field oxidation (FOX) layer 250 is then grown in the region 240 over the exposed surface of the laser recrystallized POLY layer 190. During the growth of the second FOX layer 250, dopant from the regions 173 and 174 diffuses to selected boundary limits 273 and 274, respectively, in the laser recrystallized POLY layer 190. In addition, the portion of the laser recrystallized POLY layer 190 in the region 240 oxidizes to form an oxidized POLY layer 190'.
[0026] In Figure 1 L the second nitride layer 210 and the second SRO layer 200 over the region 230 are removed to expose the laser recrystallized POLY layer 190. A positive photoresist layer 260 is the region 270 over the exposed surface of the laser recrystallized POLY layer 190 is used as a mask to form trenches 280 and 290. The trenches 280 and 290 can now be implanted with a conductive material (e.g., n+ material) to form doped POLY layers 292 and 293. The alignment of the positive photoresist layer 260 over the laser recrystallized POLY layer 190 in the region 270 is important and must be chosen such that during subsequent processing sequences the out-diffusion of both the dopant from the regions 173 and 174 as well as the implanted ions from the regions 292 and 293 will merge.
[0027] The conductive material in the trenches 280 and 290 can now be redistributed in the laser recrystallized POLY layer 190 by means of heating after removal of the positive photoresist layer 260 to form upper source and drain regions, 295 and 296 as shown in Figure 1M. During heating, redistributed conductive material in the regions 280 and 290 will meet with dopant from the regions 173 and 174 respectively, if the positive photoresist layer 260 has been properly aligned. Also, part of the POLY layer 190 will remain unaffected by the doping from the regions 295 and 296 during this heating as shown.
[0028] The structure is then completed as shown in Figure 1 N by depositing a first CVD oxide layer 300 over the entire surface of the structure, etching to form trenches 310 and 320 in the first CVD oxide layer 300, depositing a metal layer 330 over the entire surface of the structure, thereby filling in the regions 310 and 320, and selectively etching the metal layer 330 to form regions 340 and 350. A final passivation layer 360 is formed over the entire surface of the structure by CVD techniques. The passivation layer 360 can then be selectively etched leaving regions 370, 380 and 390. Contact to the underlying structure can be made through the regions 310 and 320. This fabricating method can be used to produce a structure having source and drain regions 295 and 296 of an upper transistor 500 being of n type material and source and drain regions 170 and 171 of a lower transistor 510 being of p type material with the p+ doped region 160 serving as a common gate contact between the oxide regions 180 and 150 which can be used as a complimentary MOS inverter device schematically shown in Figure 2.
[0029] Figure 3 illustrates another structure using the method in Figures 1A to 1N and shows how contact is made to the lower source and drain regions 170 and 171. This is accomplished by modification of the mask used in Figure 1Aso that the region 171 extends beyond the region 296 and across the oxidized POLY layer 190'. The oxidized POLY layer 190' in region 315 provides isolation for contact 325 in region 327.
[0030] Figure 4 shows a second embodiment of the present invention. The initial steps in Figure 4 are similar to steps A to N in Figure 1. However, in this second embodiment the first GOX layer 150 is etched over a selected region 400 leaving a surface portion of the underlying silicon substrate exposed. Subsequent deposition and doping for example with + material of a POLY layer 410 fills in this region 400. This constitutes a buried contact to the region 171.
[0031] Figure 5 illustrates a third embodiment of the present invention. The initial steps in Figure 5 are similar to the processes of Figures 3 and 4. Additionally, a layer of CVD oxide 420 is deposited over a doped POLY layer 415. An opening 425 is etched through the CVD oxide layer 420 exposing the surface of doped POLY layer 415 and a metal layer 430 is the deposited thereby filling the previously etched opening 425.
[0032] Figure 6 shows a fourth embodiment of the present invention. The steps in Figure 6 are nearly the same as in Figures 4 and 5, with the addition that four transistors 505, 510, 515, and 520 are formed at one time between the regions of first field oxidation 140 and oxidized poly 190' and all of the source and drain regions 525-530 are formed from the same type dopant (e.g., n+ type material) which is opposite in polarity to the doping of the substrate 100. The channel regions 521 and 532 associated with transistors 505 and 510 are doped to produce enhancement mode transistors (e.g., p- doping), while channel regions 535 and 536 associated with transistors 515 and 520 are implanted (e.g., n type) to produce depletion mode transistors. Polysilicon gate 540 (e.g., n+ material) is thus a common gate for enhancement mode transistors 505 and 510, while polysilicon gate 545 (e.g., n+ material) is a common gate for depletion mode transistors 515 and 520. Region 527 and polysilicon gate 545 are connected together electrically by a metal deposition (not shown) and metal contacts 550, 552, and 554 are isolated as before by oxidation layers 250 and 560 and covered by passivation layer 360. The result is a vertically stacked gate N-MOS inverter as shown schematically in Figure 7.
forming a first layer (130,150), made of a first insulating material, over a portion of the surface of the semiconductor substrate (100);
forming a second layer (160) of a first semiconductor material having doping of a second type over selected portions of said first layer (130,150);
forming first and second implant regions (170,171) of dopant of the second type in selected regions of the substrate other than the regions overlain by said second layer;
forming a third layer (172) of a second insulating material containing dopant of a first type, and subsequently removing parts of said third layer for forming a region (173,174) of the doped insulating material adjacent said second layer;
forming a fourth layer (180) of said first insulating material on the exposed portion of said second layer;
forming a fifth layer (190) of semiconductor material having doping of the second type over said first, third and fourth layers;
diffusing dopant of the first type from said third layerto a selected boundary limit in said fifth layer;
depositing dopant of the first type in selected regions of said fifth layer to within a selected lateral distance from said selected boundary limit of said second layer; and
diffusing said dopant of the first type deposited in selected regions of said fifth layer unitl said first type dopant material meets said selected boundary limit in said fifth layer.
depositing a first conductive material in selected active regions of said semiconductor substrate;
covering selected regions of said first layer (130) with a first mask (120);
removing said first layer (130) not covered by said first mask over at least one of said active regions to form an opening through said first layer, thereby exposing said semiconductor substrate surface;
removing said first mask (120); and
forming a second conductive material (160) in said opening through said first layer (130).
forming a layer (200,300) comprising insulating material on the surface of said fifth layer (190);
covering selectively a region of said layer (200,300) material with a second mask (210);
removing said layer (200,300) not covered by said second mask to form an opening (320) through said third intermediate insulating layer;
removing said second mask; and
forming a third conductive material (330) in said opening.
forming a first intermediate insulating layer over a portion of the surface of the semiconductor substrate (100);
depositing a first depletion implant of dopant of the second type in the substrate under selected regions of said first intermediate insulating layer;
forming a second layer (160) of a semiconductor material having doping of the second type over said first depletion implant;
forming first and second implant regions (170,171) of dopant of the second type in selected regions of the substrate other than the regions overlain by second layer;
forming a third layer (172) of a second intermediate insulating material containing dopant of the second type and subsequently removing parts of said third layer for forming a region (173,174) of the doped insulating material adjacent said second layer;
forming a fourth layer (180) of said first intermediate insulating layer material on the exposed portion of said second layer of semiconductor material;
forming a fifth layer (190) of semiconductor material having doping of the first type over said first, third and fourth intermediate insulating layers;
diffusing dopant of the second type from said third layer of second intermediate insulating layer to a selected boundary limit in said fifth layer;
depositing a second depletion implant of dopant of the second type in a portion of said fifth layer overlaying the selected portion of the second layer;
depositing dopant of the second type in selected regions of said fifth layer to within a selected lateral distance from said selected boundary limit of said second layer; and
diffusing said dopant of the second type deposited in selected regions of said fifth layer until said second type dopant material meets said selected boundary limit in said fifth layer.
-Ausbilden einer ersten Schicht (130; 150) eines ersten isolierenden Materials über einem Teil der Oberfläche des Halbleitersubstrats (100);
-Ausbilden einer zweiten Schicht (160) eines ersten Halbleitermaterials mit einer Dotierung eines zweiten Typs über ausgewählten Teilbereichen der ersten Schicht (130; 150);
-Ausbilden eines ersten (170) und zweiten (171) Implantiergebietes mit einer Dotierung des zweiten Typs in ausgewählten Gebieten des Substrats, die nicht von der zweiten Schicht überdeckt werden;
-Ausbilden einer dritten Schricht (172) eines zweiten isolierenden Materials mit einer Dotierung des ersten Typs und nachfolgendes Entfernen von Teilen der dritten Schicht, so daß neben der zweiten Schicht ein Gebiet (173,174) des dotierten isolierenden Materials entsteht;
-Ausbilden einer vierten Schicht (180) des ersten isolierenden Materials auf dem freiliegenden Teil der Zwuiten Schicht;
-Ausbilden einer fünften Schicht (190) von Halbleitermaterial mit einer Dotierung des zweiten Typs über der ersten, dritten und vierten Schicht;
-Diffundieren der Dotierung des ersten Typs von der dritten Schicht zu. einer ausgewählten Grenzschicht in der fünften Schicht;
-Einbringen einer Dotierung des ersten Typs in ausgewählte Gebiete der fünften Schicht in einem ausgewählten seitlichen Abstand von Rand der zweiten Schicht; und
-Diffundieren der in ausgewählte Gebiete der fünften Schicht eingebrachten Dotierung des ersten Typs, bis das Dotiermaterial des ersten Typs and die ausgewählte Grenzschicht in der fünften Schicht gelangt.
-Einbringen eines ersten Leitermaterials in ausgewählte aktive Gebiete des Halbleitersubstrats;
-Abdecken ausgewählter Gebiete der ersten Schicht (130) mit einer ersten Maske (120);
-Entfernen der nicht von der ersten Maske abgedeckten Teile de ersten Schicht (130) über wenigstens einem der aktiven Gebiete, um eine Öffnung in der ersten Schicht auszubilden, wodurch die Oberfläche des Halbleitersubstrats freigelegt wird;
-Entfernen der ersten Maske (120); und
-Ausbilden eines zweiten Leitermaterials (160) in der Öffnung in der ersten Schicht (130).
-Ausbilden einer isolierendes Material enthaltenden Schicht (200,300) auf der Oberfläche der fünften Schicht (190);
-Teilweises Abdecken eines Gebietes der Materialschicht (200,300) mit einer zweiten Maske (210);
-Entfernen der nicht von der zweiten Maske abgedeckten Teile der Schicht (200, 300), um eine Öffnung (320) in der dritten Zwischenisolierschicht auszubilden;
-Entfernen der zweiten Maske; und
-Ausbilden eines dritten Leitermaterials (330) in der Öffnung.
-Ausbilden einer erstenZwischenisol ierschicht über einem Teil der Oberfläche des Halbleitersubstrats (100);
-Einbringen eines ersten Sperrimplantats einer Dotierung eines zweiten Typs in das Substrat unterhalb ausgewählter Gebiete der ersten Zwischenisolierschicht;
-Ausbilden einer zweiten Schicht (160) eines Halbleitermaterials mit einer Dotierung des zweiten Typs über dem ersten Sperrimplantat;
-Ausbilden eines ersten (170) und zweiten (171) Implantiergebietes mit einer Dotierung des zweiten Typs in ausgewählten Gebieten des Substrats, die nicht von der zweiten Schicht überdeckt werden;
-Ausbilden einer dritten Schicht (172) eines Zwischenisoliermaterials mit einer Dotierung des zweiten Typs und nachfolgendes Entfernen Von Teilen der dritten Schicht, so daß neben der zweiten Schicht ein Gebiet (173,174) des dotierten Isoliermaterials entsteht;
-Ausbilden einer vierten Schicht (180) des ersten Zwischenisoliermaterials auf dem freiliegenden Teil derzweiten Schicht des Halbleitermaterials;
-Ausbilden einer fünften Schicht (190) von Halbleiter material mit einer Dotierung des ersten Typs über der ersten, dritten und vierten Zwischenisolierschicht;
-Diffundieren einer Dotierung des zweiten Typs von der dritten Schicht der zweiten Zwischenisolierschicht bis zu einer Grenzschicht in der fünften Schicht;
-Einbringen ecneszweiten Sperrimplantats mit einer Dotierung des zweiten Typs in as Gebiet der fünften Schicht, das den ausgewählten Teil der zweiten Schicht überdeckt;
-Einbringen einer Dotierung des zweiten Typs in ausgewählte Gebiete der fünften Schicht in einem ausgewählten seitlichen Abstand vom Rand der zweiten Schicht; und
-Diffundieren der in ausgewählte Gebiete der fünften Schicht eingebrachten Dotierung des zweiten Typs, bis das Dotiermaterial des zweiten Typs an die Ausgewählte Grenzschicht in der fünften Schicht gelangt.
-formation d'une première couche (130,150), constituée d'un matériau isolant, sur une partie de la surface du substrat semiconducteur (100),
-formation d'une seconde couche (160) en un premier matériau semiconducteur ayant un dopage d'un second type sur des parties déterminées de la première couche (130,150),
-formation d'une première et d'une seconde région implantée (170,171 ) d'un dopant du second type dans des régions déterminées du substrat, différentes des régions recouvertes par la seconde couche,
-formation d'un troisème couche (172) en un second matériau isolant contenant un dopant d'un premier type, puis retrait partiel de la troisième couche pour former une région (173,174), adjacente à la seconde couche, de matériau isolant dopé,
-formation d'une quatrième couche (180) dudit premier matériau isolant sur la partie exposée de la seconde couche,
-formation d'une cinquième couche (190) de matériau semiconducteur ayant un dopage du second type sur la première, la troisième et la quatrième couche,
-diffusion d'un dopant du premier type depuis la troisième couche vers une région frontière déterminée de la cinquième couche,
-dépôt d'un dopant du premier type dans des régions déterminées de la cinquième couche, dans les limites d'une distance latérale prédeterminée, à parit de ladite région frontière détermine de la seconde couche, et
-diffusion du dopant du premier type déposé dans des régions déterminées de la cinquième couche jusque'à ce que le matériau dopant du premier type atteigne la région limite déterminée de la cinquième couche.
-dépôt d'un premier matériau conducteur dans des régions actives prédéterminées du substrat semiconducteur,
-recouvrement des régions prédéterminées de la première couche (130) par un premier masque (120),
-retrait de la première couche (130) là où elle n'est pas recouverte par le premier masque, au moins sur l'une desdites régions actives, afin de former une ouverture dans le première couche, exposant ainsi la surface du substrate semiconducteur,
-retrait du premier masque (120), et
-formation d'un second matériau conducteur (160) dans l'ouverture pratiquée dans la première couche (130).
-formation d'une couche (200,300) comportant un matériau isolant sur la surface de la cinquième couche (190),
-recouvrement sélectif d'une région du matériau de cette couche (200,300) par un second masque,
-retrait de cette couche (200,300) là où elle n'est pas recouverte par le second masque, afin de former une ouverture (320) dans la troisième couche isolante intermédiaire,
-retrait du second masque, et
-formation d'un troisième matériau conducteur (330) dans l'ouverture.
-formation d'une première couche isolante intermediaire sur une partie de la surface du substate semiconducteur (100),
-dépôt d'un premier implant de déplétion d'un dopant du second type dans le substrat, sous des régions déterminées de la première couche isolante intermédiaire,
-formation d'une seconde couche (160) d'un matériau semiconducteur ayant un dopage du second type, au dessus du premier implant de déplétion,
-formating d'une première et d'une seconde région d'implant (170,171) de dopant du second type dans des régions déterminées du substrat autres que les régions recouvertes par la second couche.
-formation d'une troisième couche (172) d'un second matériau isolant intermédiaire comportant un dopant du second type, puis retrait partiel de la troisième couche pour former une région (173,174) de matériau isolant dopé, adjacente à la seconde couche,
-formation d'une quatrième couche (180) du matériau de la première couche isolante intermédiaire, sur la partie exposée de la second couche de matériau semiconducteur,
-formation d'une cinquième couche (190) de matériau semiconducteur avec un dopage du premier type, au dessus de la première, de la troisième, et de la quatrième couche d'isolation intermédiaire,
-diffusion d'un dopant du second type, depuis la troisième couche de la second couche d'isolation intermédiare, vers une région frontière prédéterminée dans la cinquième couche,
-dépôt d'un second implant de déplétion d'un dopant du second type dans une partie de la cinquième couche recouvrant la partie prédéterminée de la second couche,
-dépôt d'un dopant du second type dans des régions prédéterminées de la cinquième couche, dans les limites d'une distance latérale prédéterminée, à partir de la région frontière prédéterminée de la seconde couche, et
-diffusion du dopant du second type déposé dans des régions prédéterminées de la cinquième couche, jusqu'à ce que le matériau de ce dopant du second type atteigne la région frontière prédéterminée de la cinquième couche.